Artificial Intelligence is getting better than humans at detecting breast cancer
Since the early 2000s, AI systems have eliminated more than 1.7 million jobs, and that number will only increase as AI improves. Some research estimates that by 2025, AI will eliminate more than 85 million jobs.
But for all the talk about job security, AI is also proving to be a powerful tool in healthcare—specifically, cancer detection. One recently published study has shown that, remarkably, artificial intelligence was able to detect 20 percent more cancers in imaging scans than radiologists alone.
Published in The Lancet Oncology, the study analyzed the scans of 80,000 Swedish women with a moderate hereditary risk of breast cancer who had undergone a mammogram between April 2021 and July 2022. Half of these scans were read by AI and then a radiologist to double-check the findings. The second group of scans was read by two researchers without the help of AI. (Currently, the standard of care across Europe is to have two radiologists analyze a scan before diagnosing a patient with breast cancer.)
The study showed that the AI group detected cancer in 6 out of every 1,000 scans, while the radiologists detected cancer in 5 per 1,000 scans. In other words, AI found 20 percent more cancers than the highly-trained radiologists.
Scientists have been using MRI images (like the ones pictured here) to train artificial intelligence to detect cancers earlier and with more accuracy. Here, MIT's AI system, MIRAI, looks for patterns in a patient's mammograms to detect breast cancer earlier than ever before. news.mit.edu
But even though the AI was better able to pinpoint cancer on an image, it doesn’t mean radiologists will soon be out of a job. Dr. Laura Heacock, a breast radiologist at NYU, said in an interview with CNN that radiologists do much more than simply screening mammograms, and that even well-trained technology can make errors. “These tools work best when paired with highly-trained radiologists who make the final call on your mammogram. Think of it as a tool like a stethoscope for a cardiologist.”
AI is still an emerging technology, but more and more doctors are using them to detect different cancers. For example, researchers at MIT have developed a program called MIRAI, which looks at patterns in patient mammograms across a series of scans and uses an algorithm to model a patient's risk of developing breast cancer over time. The program was "trained" with more than 200,000 breast imaging scans from Massachusetts General Hospital and has been tested on over 100,000 women in different hospitals across the world. According to MIT, MIRAI "has been shown to be more accurate in predicting the risk for developing breast cancer in the short term (over a 3-year period) compared to traditional tools." It has also been able to detect breast cancer up to five years before a patient receives a diagnosis.
The challenges for cancer-detecting AI tools now is not just accuracy. AI tools are also being challenged to perform consistently well across different ages, races, and breast density profiles, particularly given the increased risks that different women face. For example, Black women are 42 percent more likely than white women to die from breast cancer, despite having nearly the same rates of breast cancer as white women. Recently, an FDA-approved AI device for screening breast cancer has come under fire for wrongly detecting cancer in Black patients significantly more often than white patients.
As AI technology improves, radiologists will be able to accurately scan a more diverse set of patients at a larger volume than ever before, potentially saving more lives than ever.
Why You Can’t Blame Your Behavior On Your Gut Microbiome
See a hot pizza sitting on a table. Count the missing pieces: three. They tasted delicious and yes, you've eaten enough—but you're still eyeing a fourth piece. Do you reach out and take it, or not?
"The difficulty comes in translating the animal data into the human situation."
Your behavior in that next moment is anything but simple: as far as scientists can tell, it comes down to a complex confluence of circumstances, genes, and personality characteristics. And the latest proposed addition to this list is the gut microbiome—the community of microorganisms, including bacteria, archaea, fungi, and viruses—that are full-time residents of your digestive tract.
It is entirely plausible that your gut microbiome might influence your behavior, scientists say: a well-known communication channel, called the gut-brain axis, runs both ways between your brain and your digestive tract. Gut bugs, which are close to the action, could amplify or dampen the messages, thereby shaping how you act. Messages about food-related behaviors could be particularly susceptible to interception by these microorganisms.
Perhaps it's convenient to imagine your resident microbes sitting greedily in your gut, crying for more pizza and tricking your brain into getting them what they want. The problem is, there's a distinct lack of scientific support for this actually happening in humans.
John Bienenstock, professor of pathology and molecular medicine at McMaster University (Canada), has worked on the gut microbiome-behavior connection for several decades. "There's a lot of evidence now in animals—particularly in mice," he says.
Indeed, his group and others have shown that, by eliminating or altering gut bugs, they can make mice exhibit different social behaviors or respond more coolly to stress; they can even make a shy mouse turn brave. But Bienenstock cautions: "The difficulty comes in translating the animal data into the human situation."
Animal behaviors are worlds apart from what we do on a daily basis—from brushing our teeth to navigating complex social situations.
Not that it's an easy task to figure out which aspects of animal research are relevant to people in everyday life. Animal behaviors are worlds apart from what we do on a daily basis—from brushing our teeth to navigating complex social situations.
Elaine Hsiao, assistant professor of integrative biology and physiology at UCLA, has also looked closely at the microbiome-gut-brain axis in mice and pondered how to translate the results into humans. She says, "Both the microbiome and behavior vary substantially [from person to person] and can be strongly influenced by environmental factors—which makes it difficult to run a well-controlled study on effects of the microbiome on human behavior."
She adds, "Human behaviors are very complex and the metrics used to quantify behavior are often not precise enough to derive clear interpretations." So the challenge is not only to figure out what people actually do, but also to give those actions numerical codes that allow them to be compared against other actions.
Hsiao and colleagues are nevertheless attempting to make connections: building on some animal research, their recent study found a three-way association in humans between molecules produced by their gut bacteria (that is, indole metabolites), the connectedness of different brain regions as measured through functional magnetic resonance imaging, and measures of behavior: questionnaires assessing food addiction and anxiety.
Meanwhile, other studies have found it may be possible to change a person's behavior through either probiotics or gut-localized antibiotics. Several probiotics even show promise for altering behavior in clinical conditions like depression. Yet how these phenomena occur is still unknown and, overall, scientists lack solid evidence on how bugs control behavior.
Bienenstock, however, is one of many continuing to investigate. He says, "Some of these observations are very striking. They're so striking that clearly something's up."
He says that after identifying a behavior-changing bug, or set of bugs, in mice: "The obvious next thing is: How [is it] occurring? Why is it occurring? What are the molecules involved?" Bienenstock favors the approach of nailing down a mechanism in animal models before starting to investigate its relevance to humans.
He explains, "[This preclinical work] should allow us to identify either target molecules or target pathways, which then can be translated."
Bienenstock also acknowledges the 'hype' that appears to surround this particular field of study. Despite the decidedly slow emergence of data linking the microbiome to human behavior, scientific reviews have appeared in brain-related scientific journals—for instance, Trends in Cognitive Sciences; CNS Drugs—with remarkable frequency. Not only this, but popular books and media articles have given the idea wings.
It might be compelling to blame our microbiomes for behaviors we don't prefer or can't explain—like reaching for another slice of pizza. But until the scientific observations yield stronger results, we still lack proof that we're doing what we do—or eating what we eat—exclusively at the behest of our resident microorganisms.
Who’s Responsible If a Scientist’s Work Is Used for Harm?
Are scientists morally responsible for the uses of their work? To some extent, yes. Scientists are responsible for both the uses that they intend with their work and for some of the uses they don't intend. This is because scientists bear the same moral responsibilities that we all bear, and we are all responsible for the ends we intend to help bring about and for some (but not all) of those we don't.
To not think about plausible unintended effects is to be negligent -- and to recognize, but do nothing about, such effects is to be reckless.
It should be obvious that the intended outcomes of our work are within our sphere of moral responsibility. If a scientist intends to help alleviate hunger (by, for example, breeding new drought-resistant crop strains), and they succeed in that goal, they are morally responsible for that success, and we would praise them accordingly. If a scientist intends to produce a new weapon of mass destruction (by, for example, developing a lethal strain of a virus), and they are unfortunately successful, they are morally responsible for that as well, and we would blame them accordingly. Intention matters a great deal, and we are most praised or blamed for what we intend to accomplish with our work.
But we are responsible for more than just the intended outcomes of our choices. We are also responsible for unintended but readily foreseeable uses of our work. This is in part because we are all responsible for thinking not just about what we intend, but also what else might follow from our chosen course of action. In cases where severe and egregious harms are plausible, we should act in ways that strive to prevent such outcomes. To not think about plausible unintended effects is to be negligent -- and to recognize, but do nothing about, such effects is to be reckless. To be negligent or reckless is to be morally irresponsible, and thus blameworthy. Each of us should think beyond what we intend to do, reflecting carefully on what our course of action could entail, and adjusting our choices accordingly.
It is this area, of unintended but readily foreseeable (and plausible) impacts, that often creates the most difficulty for scientists. Many scientists can become so focused on their work (which is often demanding) and so focused on achieving their intended goals, that they fail to stop and think about other possible implications.
Debates over "dual-use" research exemplify these concerns, where harmful potential uses of research might mean the work should not be pursued, or the full publication of results should be curtailed. When researchers perform gain-of-function research, pushing viruses to become more transmissible or more deadly, it is clear how dangerous such work could be in the wrong hands. In these cases, it is not enough to simply claim that such uses were not intended and that it is someone else's job to ensure that the materials remain secure. We know securing infectious materials can be error-prone (recall events at the CDC and the FDA).
In some areas of research, scientists are already worrying about the unintended possible downsides of their work.
Further, securing viral strains does nothing to secure the knowledge that could allow for reproducing the viral strain (particularly when the methodologies and/or genetic sequences are published after the fact, as was the case for H5N1 and horsepox). It is, in fact, the researcher's moral responsibility to be concerned not just about the biosafety controls in their own labs, but also which projects should be pursued (Will the gain in knowledge be worth the possible downsides?) and which results should be published (Will a result make it easier for a malicious actor to deploy a new bioweapon?).
We have not yet had (to my knowledge) a use of gain-of-function research to harm people. If that does happen, those who actually released the virus on the public will be most blameworthy–-intentions do matter. But the scientists who developed the knowledge deployed by the malicious actors may also be held blameworthy, especially if the malicious use was easy to foresee, even if it was not pleasant to think about.
In some areas of research, scientists are already worrying about the unintended possible downsides of their work. Scientists investigating gene drives have thought beyond the immediate desired benefits of their work (e.g. reducing invasive species populations) and considered the possible spread of gene drives to untargeted populations. Modeling the impacts of such possibilities has led some researchers to pull back from particular deployment possibilities. It is precisely such thinking through both the intended and unintended possible outcomes that is needed for responsible work.
The world has gotten too small, too vulnerable for scientists to act as though they are not responsible for the uses of their work, intended or not. They must seek to ensure that, as the recent AAAS Statement on Scientific Freedom and Responsibility demands, their work is done "in the interest of humanity." This requires thinking beyond one's intentions, potentially drawing on the expertise of others, sometimes from other disciplines, to help explore implications. The need for such thinking does not guarantee good outcomes, but it will ensure that we are doing the best we can, and that is what being morally responsible is all about.